Computing
OpenAI’s Bigger Bug Bounty Puts AI Security Under the Microscope
OpenAI’s expanded AI security bug bounty offers up to $172,000 for exceptional critical vulnerabilities.
▚Computing
Computing
Gorbag Targets Ukraine With Data-Stealing Phishing Campaigns
Gorbag combines spearphishing, malicious archives and a self-deleting stealer to target Ukrainian courts, councils and defense organizations.
Computing
HTTP2Shell: Why Runtime Virtual Patching Matters Now
HTTP2Shell exposes the widening gap between AI-assisted bug discovery and production patching. Runtime virtual patching can protect exposed workloads…
Computing
AI Security Training Needs a Human Decision Layer
AI can make security information faster and easier to understand, but people must remain responsible for context, judgment and behavior…
Computing
Election Scams Target Registration, Donations and Absentee Lists
Election scams use texts, calls and email to impersonate voter services. Learn how to verify registration claims, fines and donation…
Computing
AI Speeds Up Flaws and Compresses the Patching Window
AI-driven vulnerability discovery and agentic access are compressing patch deadlines, forcing security teams to prioritize exposure, containment…
Computing
Security Reporting Fails When Staff Fear the Response
Better incident reporting starts with a response that separates bad events from blame and makes flagging suspicious activity easy.
Computing
AI Finds More Bugs, but Human Judgment Still Sets the Bounty
Automated discovery is expanding security research, but human context and impact analysis now carry greater value in reward decisions.
Computing
Customer Relationship Fraud Is Moving Beyond Onboarding
Customer relationship fraud now targets established accounts and trusted communications, making continuous identity checks and transaction controls…
Computing
AI Agent Identity Needs Owners and Access Controls
AI agents need governed identities, accountable owners, and carefully reviewed access as they work alongside employees.
Computing
AI Coding Security: When Agents Pull In Dead MCP Servers
AI coding security risks grow as agents install deprecated MCP servers with known flaws and no future patches.
Computing
Borrowed Infrastructure Phishing: When Trusted Platforms Carry the Bait
Borrowed infrastructure phishing turns real domains, certificates and login prompts into attack tools. Here’s how defenders should respond.